ADDZYME: A software to predict effect of additives on enzyme activity

Abstract

Enzymes are biological catalysts that accelerate chemical reactions by reducing their activation energy. Enzymes specific environmental conditions to function optimally. Additive molecules and compounds, such as organic solvents, ionic liquids, and deep eutectic solvents, have crucial effects on enzyme behavior by changing activity and stability. However, finding and testing different additives is an expensive process that requires specialists, laboratory equipment, and chemical compounds. Machine learning, which has been present in all fields of science and technology in recent years, is one of the ways to find a suitable additive for our desired enzyme without doing a time-consuming experimental process. In this manuscript, we introduce ADDZYME, a machine learning-based software, to predict the effect of additives on enzyme activity. ADDZYME utilizes an ensemble of extremely randomized trees models alongside physicochemical descriptors to make a prediction. To ease usage, ADDZYME is accompanied by a graphical user interface. ADDZYME is freely available on www.github.com/miladrayka/addzyme for further experiments.

Graphical abstract

SYNOPSIS: ADDZYME software applies a machine-learning algorithm to predict the effect of an additive on an enzyme activity. ADDZYME utilizes an ensemble of extremely randomized trees models alongside physicochemical descriptors to make a prediction.